Cortical dysfunction has been implicated in many neurological disorders including epilepsy, schizophrenia, and stroke. The majority of cortical connections are local, implying that local microcircuitry should be a dominant contributor to pyramidal neuron receptive fields and, therefore, perception. Cortical microcircuit motifs include recurrent excitation and inhibition, which mediate interactions between similarly tuned populations, and feed- forward excitation and lateral inhibition, which mediate interactions between distinctly tuned populations. Studying these motifs in vivo has proven challenging because pyramidal neurons with different functional tuning are often intermingled, requiring cellular-resolution perturbation approaches to probe microcircuit function. The objective of this proposal is to test the hypothesis that local microcircuit interactions shape neural receptive fields during naturalistic behavior, and that these interactions contribute to perception. We focus on mouse primary vibrissal somatosensory cortex (vS1), where sensory input from single whiskers outputs onto small patches of cortex known as `barrels', making individual whisker sensory representations tractable targets to comprehensive recording and subsequent perturbation. In previously published work, we demonstrated the ability to record and classify the majority of layer (L) 2/3 neurons in a barrel. Here, we present preliminary data demonstrating the ability to lesion small subsets of identified neurons in a barrel using multiphoton ablation, thereby overcoming the previous constraint on experiments probing the role of recurrent amplification among similarly tuned neurons. Preliminary experiments in single-whisker mice indicate that recurrent excitation in vS1 L2/3 amplifies the responses of neurons tuned to whisker touch, but not of those tuned to whisker movement, and that recurrent inhibition does not exert a measurable effect on touch responses. Further preliminary data in mice with two whiskers reveals that feed-forward excitation from single-whisker responsive neurons shapes multi-whisker responses and that cross-whisker suppression declines following single-whisker neuron lesions. Finally, preliminary barrel-scale lesion experiments reconcile recent controversies in the field and show that vS1 is necessary for discrimination but not detection behaviors. We propose three aims testing 1) whether recurrent interactions ? excitatory and inhibitory ? shape vS1 L2/3 responses of neurons tuned to the same whisker; 2) whether L2/3 excitatory touch neurons tuned to different whiskers interact via feed-forward excitation to generate multi-whisker receptive fields and via lateral inhibition to produce cross-whisker suppression; 3) whether individual vS1 barrels contribute to perception, and whether L2/3 recurrent excitation in vS1 contributes to perception. The proposed work involves a novel combination of large-scale two-photon calcium imaging, multiphoton ablation, barrel-scale lesions, and quantitative head-fixed mouse behavior. Our long-term goal is to understand sensory microcircuit computations and how they shape perception.